Effects of Cognitive Task Type and Complexity on Dual-Task Interference During Level-Ground Walking and Obstacle Negotiation in Individuals with Stroke

Abstract

Background: Compromised dual-task walking ability reduces functional independence in community-dwelling individuals after stroke. Objective. To examine the influence of mobility task and cognitive task type and complexity, and their interaction on dual-task level-ground walking and obstacle-crossing after stroke. Methods: Ninety-three individuals with chronic stroke (mean [SD] age = 62.4 [6.7] years, stroke duration = 67.7 [53.5] months) participated in this observational study with repeated measures. For each dual-task testing condition, a mobility task (level-ground walking or obstacle-crossing) was performed concurrently with 1 of 5 cognitive tasks (serial-subtractions, category naming, clock test, auditory discrimination, and shopping-list recall). Each cognitive task involved low and high complexity levels, yielding 20 dual-task conditions in total. Dual-task effect (DTE = [single-task − dual-task]×100/single-task) on walking distance (mobility-DTE) and number of correct responses (cognitive-DTE) were calculated for each dual-task condition. Results: Medium to large interaction effects were observed between cognitive task type and complexity on cognitive (F = 12.0-15.8, P < .001, ηp2 = 0.12-0.15) and mobility performance (F = 3.2-5.5, P < .05, ηp2 = 0.03-0.06) during dual-task level-ground walking and obstacle-crossing. Among the cognitive tasks, serial-subtraction had the greatest interference effect on both cognitive (Mean DTE = −9.2 to −21.5%) and mobility performance (Mean DTE = −18.7 to −19.1%). Overall, “mobility interference” (decrement in walking distance without a decrement in cognitive performance) was the most common dual-task effect pattern observed. Conclusion: The type and complexity level of the mobility and cognitive tasks interact to influence the degree and pattern of dual-task effects, with the serial-subtraction task inducing the greatest effect. Standardized assessments involving distinct cognitive domains are necessary for profiling dual-task interference during walking among individuals with chronic stroke.